Using a boundary layer technique to reduce chatter in sliding mode controllers

A changing boundary layer width is used to adjust for the tradeoff between robustness and chatter elimination. Speed control of a DC motor is used as an example to demonstrate the effectiveness of the proposed approach. The boundary layer width is adjusted from a very large value (to eliminate chatter in the transient period) to a small value (to improve robustness as the system approaches steady state). A design model of the DC motor and a sliding mode controller using saturation type functions in place of the switching functions are developed. A discussion about the width of the boundary layer, chatter, robustness, and the effect of small time delays is presented. Performance with fixed boundary layer widths is simulated for various operating conditions in the motor speed example. A function of the state is then chosen to fit the gains, and response to the proposed controller is simulated. Comparisons to another boundary layer adjustment technique are drawn.<
>